Abstract
Data obfuscation is a well-known technique for protecting user privacy against inference attacks, and it was studied in diverse settings, including search queries, recommender systems, location-based services and Online Social Networks (OSNs). However, these studies typically take the point of view of a single user who applies obfuscation, and focus on protection of a single target attribute. Unfortunately, while narrowing the scope simplifies the problem, it overlooks some significant challenges that effective obfuscation would need to address in a more realistic setting. First, correlations between attributes imply that obfuscation conducted to protect a certain attribute, may influence inference attacks targeted at other attributes. In addition, when multiple users conduct obfuscation simultaneously, the combined effect of their obfuscations may be significant enough to affect the inference mechanism to their detriment. In this work we focus on the OSN setting and use a dataset of 1.9 million Facebook profiles to demonstrate the severity of these problems and explore possible solutions. For example, we show that an obfuscation policy that would limit the accuracy of inference to 45% when applied by a single user, would result in an inference accuracy of 75% when applied by 10% of the users. We show that a dynamic policy, which is continuously adjusted to the most recent data in the OSN, may mitigate this problem. Finally, we report the results of a user study, which indicates that users are more willing to obfuscate their profiles using popular and high quality items. Accordingly, we propose and evaluate an obfuscation strategy that satisfies both user needs and privacy protection.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Ardagna, C.A., Cremonini, M., Damiani, E., De Capitani di Vimercati, S., Samarati, P.: Location privacy protection through obfuscation-based techniques. In: Barker, S., Ahn, G.-J. (eds.) Data and Applications Security 2007. LNCS, vol. 4602, pp. 47–60. Springer, Heidelberg (2007)
Breiman, L.: Random forests. Machine Learning 45(1), 5–32 (2001)
Chaabane, A., Acs, G., Kaafar, M.A.: You are what you like! Information leakage through users’ Interests. In: Proceedings of the 19th NDSS (2012)
Chen, T., Boreli, R., Kaafar, M.-A., Friedman, A.: On the effectiveness of obfuscation techniques in online social networks. Technical Report 1833-9646-8065, National ICT Australia (April 2014)
du Pin Calmon, F., Fawaz, N.: Privacy against statistical inference. In: Allerton Conference, pp. 1401–1408 (2012)
Guyon, I., Elisseeff, A.: An introduction to variable and feature selection. J. Mach. Learn. Res. 3, 1157–1182 (2003)
He, J., Chu, W.W.: Protecting private information in online social networks. In: Chen, H., Yang, C.C. (eds.) Intelligence and Security Informatics. SCI, vol. 135, pp. 249–273. Springer, Heidelberg (2008)
He, J., Chu, W.W., Liu, Z(V.): Inferring Privacy Information from Social Networks. In: Mehrotra, S., Zeng, D.D., Chen, H., Thuraisingham, B., Wang, F.-Y. (eds.) ISI 2006. LNCS, vol. 3975, pp. 154–165. Springer, Heidelberg (2006)
Kosinski, M., Stillwell, D., Graepel, T.: Private Traits and Attributes are Predictable from Digital Records of Human Behavior. PNASÂ 110(15) (2013)
Li, C., Shirani-Mehr, H., Yang, X.: Protecting individual information against inference attacks in data publishing. In: Kotagiri, R., Radha Krishna, P., Mohania, M., Nantajeewarawat, E. (eds.) DASFAA 2007. LNCS, vol. 4443, pp. 422–433. Springer, Heidelberg (2007)
Lindamood, J., Heatherly, R., Kantarcioglu, M., Thuraisingham, B.: Inferring Private Information Using Social Network Data. In: WWW (2009)
Manning, C.D., Raghavan, P., Schütze, H.: Introduction to Information Retrieval. Cambridge University Press, New York (2008)
Mislove, A., Viswanath, B., Gummadi, K.P., Druschel, P.: You Are Who You Know: Inferring User Profiles in Online Social Networks. In: WSDM (2010)
Peddinti, S.T., Saxena, N.: On the privacy of web search based on query obfuscation: A case study of trackMeNot. In: Atallah, M.J., Hopper, N.J. (eds.) PETS 2010. LNCS, vol. 6205, pp. 19–37. Springer, Heidelberg (2010)
Ryu, E., Rong, Y., Li, J., Machanavajjhala, A.: Curso: Protect yourself from Curse of Attribute Inference: a Social Network Privacy-analyzer. In: WDSN (2013)
Salamatian, S., Zhang, A., du Pin Calmon, F., Bhamidipati, S., Fawaz, N., Kveton, B., Oliveira, P., Taft, N.: How to Hide the Elephant or the Donkey in the Room: Practical Privacy Against Statistical Inference for Large Data. In: GlobalSIP (2013)
Weinsberg, U., Bhagat, S., Ioannidis, S., Taft, N.: Blurme: inferring and obfuscating user gender based on ratings. In: RecSys. ACM (2012)
Ye, S., Wu, F., Pandey, R., Chen, H.: Noise Injection for Search Privacy Protection. In: Conference on Computational Science and Engineering (2009)
Zheng, Z., Wu, X., Srihari, R.: Feature selection for text categorization on imbalanced data. SIGKDD Explor. Newsl. 6(1), 80–89 (2004)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Chen, T., Boreli, R., Kaafar, MA., Friedman, A. (2014). On the Effectiveness of Obfuscation Techniques in Online Social Networks. In: De Cristofaro, E., Murdoch, S.J. (eds) Privacy Enhancing Technologies. PETS 2014. Lecture Notes in Computer Science, vol 8555. Springer, Cham. https://doi.org/10.1007/978-3-319-08506-7_3
Download citation
DOI: https://doi.org/10.1007/978-3-319-08506-7_3
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-08505-0
Online ISBN: 978-3-319-08506-7
eBook Packages: Computer ScienceComputer Science (R0)